Mental Health First Aid: 5 Failures That Hide Risk

HSE reports that 964,000 workers in Great Britain suffered work-related stress, depression or anxiety in 2024/25, a number large enough to expose the limits of awareness campaigns that never change work design. Mental health first aid can help a distressed worker in the moment, but this diagnostic shows the 5 failures that turn a caring program into a weak substitute for risk control.

Mental health first aid at work is a trained peer or manager response that recognizes distress, gives initial support, and routes the worker toward appropriate help. It is not a psychosocial risk management system, because ISO 45003 expects organizations to identify hazards, control work factors, consult workers, and review effectiveness.

Why is mental health first aid not enough by itself?

Mental health first aid is not enough by itself because it usually acts after distress is visible, while workplace mental-health risk often starts earlier in workload, autonomy, role conflict, bullying, fatigue, trauma exposure, or poor return-to-work decisions. ISO published ISO 45003:2021 as guidance for managing psychosocial risk within an occupational health and safety management system, which places mental health inside the same logic as hazard identification, control, participation, and review.

As Andreza Araujo argues in Safety Culture: From Theory to Practice, culture becomes visible in repeated decisions, not in declared values. A company can train 200 first aiders and still leave supervisors rewarding impossible deadlines, tolerating disrespect, or sending employees back to the same work pattern that made them ill. That is care after exposure, not control before exposure.

The practical test is simple. If a mental health first aider hears the same pattern from 3 workers in the same unit within 30 days, the program should trigger a management review, not only another confidential referral. The issue may be workload, staffing, shift design, harassment, role ambiguity, or a leadership routine that treats exhaustion as commitment.

1. Failure: treating training as the control

Training becomes a false control when leaders count certified mental health first aiders but cannot name which psychosocial hazards the organization is trying to reduce. A 2-day course may improve recognition and conversation quality, yet it does not change overtime demand, job control, conflict escalation, or work pacing unless the organization connects the course to risk governance.

Across 25+ years leading EHS at multinationals, Andreza Araujo has observed the same pattern in safety programs that look mature on paper. Training is visible, easy to schedule, and easy to report, while redesigning work asks leaders to challenge planning, staffing, budget, and authority. Mental health first aid falls into the same trap when completion rates become the headline metric.

Use training as an entry point, not as proof of prevention. The EHS or HR owner should define 3 escalation triggers: repeated concerns from one team, a distress event linked to a work factor, and a return-to-work case that cannot be accommodated without changing workload. That keeps mental health first aid connected to the workplace mental health escalation protocol instead of isolating it as a good-intentioned course.

2. Failure: hiding workload risk behind individual resilience

Workload risk hides when mental health first aid conversations frame distress as an individual coping problem before testing the design of the job. WHO states that almost 60% of the global population is in work, which means work organization is not a side issue in mental health. It is one of the main environments where risk can be created or reduced.

What most mental health campaigns understate is the moral hazard of resilience language. When a worker is overloaded by chronic understaffing, unclear priorities, and constant interruption, resilience training can become a softer vocabulary for tolerating poor design. Andreza Araujo's critique of compliance theater applies here: the organization may appear supportive while its operating rhythm keeps producing harm.

Leaders should convert first-aid signals into a workload review within 14 days when the same unit shows repeated distress, absence, errors, or conflict. The review should compare demand, control, support, role clarity, change load, and recovery time. The method aligns naturally with a workload risk triage, because the question is no longer whether the worker can cope. The question is whether the work is being managed within human limits.

3. Failure: confusing referral with prevention

Referral is necessary when a worker needs clinical, employee assistance, or emergency support, but referral does not prove that the workplace has prevented recurrence. WHO and ILO issued 2022 guidance that covers organizational interventions, manager training, worker training, individual interventions, return to work, and employment support, which means a credible program must move beyond signposting.

During the PepsiCo South America tenure, where the accident ratio fell 50% in 6 months, Andreza Araujo learned that follow-up changed results more than slogans did. The same logic applies to mental health at work. A referral without feedback on work causes, within confidentiality limits, leaves the organization blind to the exposure it may be repeating.

Build a closed loop that protects privacy while still learning from patterns. Track anonymous themes, affected work groups, role factors, shift patterns, manager involvement, and return-to-work barriers. This is where many EAP programs fail, because they count use of service as activity while never asking whether the organization reduced the hazard. The article on EAP program traps expands that distinction.

4. Failure: leaving supervisors outside the system

Supervisors become the missing control when they are told to be supportive but are not given decision rights, escalation routes, or workload authority. A first aider may identify distress, yet the supervisor often controls the schedule, job allocation, conflict handling, return-to-work adjustments, and the daily signals that decide whether the worker feels safe to speak.

In Make The Difference: Be a Leader in Health & Safety, Andreza Araujo treats leadership as observable action under pressure. For mental health first aid, that means the supervisor cannot remain a spectator who says the right words while the work system stays unchanged. The supervisor must know what can be adjusted today, what requires HR approval, and what must be escalated within 24 hours.

Give supervisors a 5-question field script: what changed in the work, what demand is now unrealistic, what support is missing, what decision is blocked, and what temporary adjustment can reduce exposure this week. The script does not turn supervisors into clinicians. It turns them into leaders who can remove work factors before the situation becomes a case file.

5. Failure: measuring activity instead of risk reduction

Activity metrics distort mental health first aid when dashboards celebrate trained headcount, awareness sessions, posters, and referral volume without testing whether psychosocial risk declined. A useful dashboard should show both support capacity and risk movement, because an increase in referrals may indicate trust, hidden exposure, or deterioration, depending on the context.

In more than 250 cultural transformation projects supported by Andreza Araujo's team, the stronger question is rarely "how much activity happened?" The stronger question is "which decision changed because the evidence became visible?" Mental health metrics should pass the same test. If no leader changes staffing, pacing, role clarity, conflict response, or return-to-work practice, the dashboard is describing motion.

Track 6 indicators together: trained first aiders per site, repeated themes by team, time from concern to management review, workload actions closed, return-to-work adjustments sustained after 30 days, and recurrence of distress signals in the same work group. That mix keeps care, control, and learning in the same frame.

What should replace the training-only model?

A better model keeps mental health first aid as one layer in a 4-layer system: recognition, escalation, risk control, and review. Recognition identifies distress, escalation routes urgent and repeated patterns, risk control changes work factors, and review tests whether the change held after 30, 60, and 90 days.

The practical owner should be shared, not blurred. HR usually owns clinical support pathways, privacy, accommodation, and employee relations. EHS should own psychosocial hazard assessment, risk controls, and management-system integration. Operations must own workload, staffing, priorities, and supervisor behavior. The failure starts when everyone supports mental health but nobody owns the work factors.

Use a decision-rights matrix for psychosocial risk, especially in multi-site organizations where local managers can normalize overload. The companion article on a psychosocial decision-rights matrix shows how to separate consultation from authority, which is essential when privacy, discipline, workload, and safety overlap.

How do you know the program is becoming a liability?

The program becomes a liability when the organization can prove it trained people to recognize distress but cannot prove it acted on repeated work-related signals. That gap matters because a mature employer should not only respond kindly to suffering. It should also ask whether its own management system created or prolonged the exposure.

The warning signs are concrete. The same department produces repeated concerns, referrals rise while workload stays unchanged, supervisors ask HR to "handle" every case, return-to-work plans send people back to the same pressure, and dashboards report course completion without risk controls. When those 5 signs appear together, mental health first aid may be protecting the brand more than the worker.

Andreza Araujo's book A Ilusao da Conformidade ("The Illusion of Compliance") is useful for this point because mental health programs can look caring while functioning as documentation. The issue is not bad faith. The issue is a control gap that polite language can hide for months.

Each month without a control loop allows repeated distress patterns to become normalized, while the organization keeps investing in visible support that may never touch workload, leadership behavior, or return-to-work design.

Mental health first aid vs psychosocial risk control

Mental health first aid and psychosocial risk control answer different questions. First aid asks how the organization supports a person showing distress today. Risk control asks which work factors made distress more likely and what leaders must change so the exposure does not repeat.

Conclusion

Mental health first aid belongs in the workplace, but it should never become the elegant excuse for ignoring psychosocial hazards that leaders can actually control.

Safety is about coming home, including psychologically. If your organization needs to connect mental health support with psychosocial risk assessment, supervisor routines, and safety culture, Andreza Araujo and ACS Global Ventures can support a practical diagnostic through Andreza Araujo.